[dbaupp]

Concepts like finite integer sizes (aka integer overflow) and finite floating point precision (1/3+1/3+1/3 != 1) aren't PhD-level concerns. :)